Search for: All records

We propose an efficient estimator for the coefficients in censored quantile regression using the envelope model. The envelope model uses dimension reduction techniques to identify material and immaterial components in the data, and forms the estimator based only on the material component, thus reducing the variability of estimation. We will demonstrate the guaranteed asymptotic efficiency gain of our proposed envelope estimator over the traditional estimator for censored quantile regression. Our analysis begins with the local weighing approach that traditionally relies on semiparametric ‐estimation involving the conditional Kaplan–Meier estimator. We will instead invoke the independent identically distributed (i.i.d.) representation of the Kaplan–Meier estimator, which eliminates this infinite‐dimensional nuisance and transforms our objective function in ‐estimation into a ‐process indexed by only an Euclidean parameter. The modified ‐estimation problem becomes entirely parametric and hence more amenable to analysis. We will also reconsider the i.i.d. representation of the conditional Kaplan–Meier estimator.

 

Coastal aquifers supply freshwater to nearly half the global population, yet they are threatened by salinization. Salinities are typically estimated assuming steady‐state, neglecting the effect of cyclical forcings on average salinity distributions. Here, numerical modeling is used to test this assumption. Multi‐scale fluctuations in sea level (SL) are simulated, from tides to glacial cycles. Results show that high‐frequency fluctuations alter average salinities compared with the steady‐state distribution produced by average SL. Low‐frequency forcing generates discrepancies between present‐day salinities estimated with and without considering the cyclical forcing due to overshoot effects. This implies that salinities in coastal aquifers may be erroneously estimated when assuming steady‐state conditions, since present distributions are likely part of a dynamic steady state that includes forcing on multiple timescales. Further, typically neglected aquifer storage characteristics can strongly control average salinity distributions. This has important implications for managing vulnerable coastal groundwater resources and for calibration of hydrogeological models.

 

Tidal wetlands are comprised of complex interdependent pathways where measurements of carbon exchange are often scale dependent. Common data collection methods (i.e., chambers and eddy covariance) are inherently constrained to different spatial and temporal scales which could generate biased information for applications of carbon accounting, identifying functional relationships and predicting future responses to climate change. Consequently, it is needed to systematically evaluate measurements derived from multiple approaches to identify differences and how techniques complement each other to reconcile interpretations. To accomplish this, we tested ecosystem‐scale eddy covariance with plot‐scale chamber measurements within a temperate salt marsh. We found good agreement (R² = 0.71–0.95) when comparing measurements of CH₄emissions and CO₂exchange but this agreement was dependent upon canopy phenology with discrepancies mainly arising during senescence and dormancy phenophases. The environmental drivers for CH₄and CO₂fluxes were mostly preserved across different measurement techniques, but the number of drivers increases while their individual strength decreases at the ecosystem scale. Empirical upscaling models parameterized with chamber measurements overestimated annual net ecosystem exchange (NEE; 108%) and gross primary production (GPP; 12%) while underestimating ecosystem respiration (Reco; 14%) and CH₄emissions (69%) compared to eddy covariance measurements. Our results suggest that the environmental complexity of CH₄and CO₂fluxes in salt marshes may be underestimated by chamber‐based measurements, and highlights how different techniques are complementary while considering limitations at each level of measurement.

 

We conduct a stated‐preference choice experiment to reveal motorists' driving‐related behavioral responses to different types of signs indicating that the road is flooded and travel costs associated with avoidance of the flooded road. We use three flood‐indicating visualization treatments and control group to identify the effects of particular road signs and identify associations between drivers' behavior and their demographic characteristics and the cost (time) of taking an alternate route. Using responses from 714 adult participants from the coastal area of the Mid‐Atlantic of the United States, we estimate willingness to drive additional minutes to avoid flooded roads using a random utility framework. Our results suggest that individuals are more likely to avoid flooded roads when shown flood‐indicating road signs that do not indicate the exact depth of the water and signs that indicate that the water is relatively deep (more than 12 in.). We further find that individuals tend to persist in their initial choices. They often make risky choices when high risk‐indicating information is presented at the beginning of the decision‐making process. The results of this study can help inform the sign design choices of transportation managers to help ensure driver safety in flood conditions.

 

As sediment is transported through river corridors, it typically spends more time in storage than transport, and as a result, sediment delivery timescales are controlled by the duration of storage. Present understanding of storage timescales is largely derived from models or from field studies covering relatively short (≤10² year) time spans. Here we quantify the storage time distribution for a 17 km length of Powder River in Montana, USA by determining the age distribution of eroded sediment. Our approach integrates surveyed cross‐sections, analysis of historical aerial imagery, aerial LiDAR, geomorphic mapping, and age control provided by optically stimulated luminescence (OSL) and dendrochronology. Sediment eroded by Powder River from 1998 to 2013 ranges from a few years to ∼5,000 years in age; ages are exponentially distributed (r² = 0.78; Anderson‐Darlingpvalue 0.003). Eroded sediment is derived from Powder River's meander belt (∼900 m wide), which is only 1.25 times its meander wavelength, a value reflecting valley confinement rather than free meandering. The mean storage time, 824 years (95% C.I. 610–1030 years), is similar to the time required to rework deposits of Powder River's meander belt based on an average meander migration rate of ∼1 m/yr, implying that storage time distributions of confined meandering rivers can be quantified from remotely sensed estimates of meander belt width and channel migration rates. Heavy‐tailed storage time distributions, frequently cited from physical and numerical modeling studies, may be restricted to unconfined meandering rivers.

 

This paper develops a coupled hydrologic‐economic model that estimates the effects of six tax institutions that theory predicts will lead to equal amounts of aquifer withdrawal. That said, the distributive effects of the tax institutions are expected to differ because each involves different combinations of tax thresholds and side payments (returned tax revenue). The tax policies can lead to groundwater users being worse off than they would be using an unmanaged aquifer. This study explores whether the distributive impacts of specific policies that have equal marginal incentives lead to differences in the behavior of participants in an experiment involving a common pool groundwater resource. The results reveal that each of the tax policies results in approximately the same reduction in resource use but affects participants' earnings and opinions regarding the policies differently. A tax imposed on groundwater use above a threshold and without a side payment is most effective in increasing the net social benefit associated with using the aquifer; participant earnings under that scheme are almost equal to earnings from an unmanaged aquifer (1.04% less), and overall social efficiency is greater (4.34%). Unfortunately, participants tend to prefer an unmanaged aquifer and tax policies with high side‐payments—treatments that led to lower overall social efficiency. The evidence suggests that aquifer management may require a two‐fold approach: (1) a carefully selected threshold that can make water users financially indifferent between a managed and unmanaged aquifer and (2) education to increase the political acceptability of the managed aquifer policy.

 

Extreme precipitation events are arguably one of the most important natural hazards in many areas of the globe, impacting nearly every societal sector. In the Northeastern United States, extreme precipitation events have been shown to be increasing with several recent events garnering national attention (i.e., Ellicott City Maryland 2018; Tropical Storm Lee 2011). The NOAA Atlas 14 product is the nation's standard for estimating the magnitude and frequency of site‐specific extreme precipitation events, containing both precipitation frequency estimates, as well as associated confidence intervals. The Atlas uses surface stations, primarily from the National Weather Service Cooperative Observer Program, and statistical methodologies to provide point‐based precipitation exceedance probability estimates for several durations and potential recurrence intervals. Unfortunately, the number and quality of Cooperative Observer sites varies greatly over space and time. This research compares observed precipitation extremes from a high‐resolution statewide mesonet to those estimated by the Atlas 14 product for a 10‐year recurrence interval at several precipitation durations. Results of the analysis indicate that Atlas 14 underestimates the number and magnitude of extreme precipitation events across the state of Delaware at longer event durations (360‐ to 1,440‐min). At shorter durations (5‐ to 240‐min) the Atlas 14 estimates are more closely aligned with the observations from the high‐resolution precipitation network. These results suggest that caution should be exercised when using Atlas 14 estimates for engineering standards and hydrologic studies, especially for longer duration events. Therefore, a more rapid update cycle for revision of the Atlas 14 product should be considered, as a changing climate regime may be responsible for the differences identified in this research.

 

Concrete contains low microbial biomass, but some bacteria can grow in this highly alkaline environment. We used silica-based DNA extraction and 16S rRNA sequence analysis to identify the bacteria in a corroded concrete bridge sample from Bethlehem, Pennsylvania. Staphylococcus, Streptococcus, Corynebacterium, Leifsonia, Vicinamibacterales, and Actinophytocola were the most abundant genera. 

Declines in commercial crustacean species (such as lobsters, king crab, etc.) have caused an increased interest in the harvest of the red deep-sea crab Chaceon quinquedens. The red deep-sea crab is a federally managed fishery; however, little is known about the species’ general biology, especially the conditions required for larval survival. We aimed to answer two main questions about the life history of the red deep-sea crab. First, is there a common larval hatching pattern between adult female crabs? Specifically, our inquiries are about the duration of the hatching process, daily peak hatching time, and the relationship between female morphometry and the total larvae hatched. Second, which are the factors affecting the survival and development of larval red deep-sea crabs? In order to answer these research questions, we studied the effects of diet (rotifers, Artemia sp., algae, and unfed), temperature (9 °C, 15 °C, and 20 °C), and aquaculture settings. Ovigerous females were obtained from commercial traps and transported to the NOAA James J. Howard Laboratory, NJ. They were placed in the Females Husbandry and Hatching Collection System (FHCS), where the larvae hatched. Hatching of adult females was monitored and measured by volume. A simple linear regression (SLR) was calculated to predict the number of larvae hatched based on the measured volumes, and it was significant (F = 1196; df = 1, 13; R2 = 0.9892, p = 3.498 × 10−14). Duration of hatching period showed an approximate 30 days for adult females red deep-sea crabs, with a common daily maximum hatching time at 22:00 hrs (hatching time seem to follow the sun cycle and the first hours after sunset, Perez, pers. observation). Linear polynomial quadratic regressions were conducted for both years with an interaction term for the two continuous variables (diet and temperature), and were used to model the proportion of larval survival through time. In both years, a highly significant difference was obtained (F = 56.15; df = 4, 2134; R2 = 0.09353; p = < 2.2 × 10−16). There is an effect of diet and temperature in the survival of red deep-sea crabs, but not a combined effect of them. 

Economic experiments have emerged as a powerful tool for agricultural policy evaluations. In this perspective, we argue that involving stakeholders in the design of economic experiments is critical to satisfy mandates for evidence-based policies and encourage policymakers' usage of experimental results. To identify advantages and disadvantages of involving stakeholders when designing experiments, we synthesize observations from six experiments in Europe and North America. In these experiments, the primary advantage was the ability to learn within realistic decision environments and thus make relevant policy recommendations. Disadvantages include complicated implementation and constraints on treatment design. We compile 12 recommendations for researchers.